Color-image-sensing apparatuses employing a photoelectric transducer (image sensor), such as a CCD and a CMOS sensor, have been commonly utilized. In these image-sensing apparatuses, by making light that has passed through a color filter enter an image sensor, color-image sensing is realized through an image sensor incapable of discriminating color.
FIG. 11 is a block diagram illustrating an example of a signal processing circuit of an image-sensing apparatus utilizing a single-plane type image sensor having a primary color filter as illustrated in FIG. 12.
Black-balance adjustment and white-balance adjustment are applied to the output signal of an image sensor (a CCD, in this case) 1001 in an OB circuit 1002 and a WB circuit 1003, respectively. The output of a color interpolation circuit 1004 is converted by a color conversion circuit 1005 into color-difference signals (R-Y, B-Y), to which, in a CSUP (Chroma Suppress) circuit 1006, false-color reduction processing is applied with regard to saturated areas. Meanwhile, the output of the WB circuit 1003 is converted by a luminance interpolation circuit 1007 into a luminance signal (Yh), edge portions of which are then enhanced by an edge-enhancement (APC) circuit 1008. The color-difference signals from the CSUP circuit 1006 and the luminance signal from the edge-enhancement circuit 1008 are converted by a color conversion circuit 1009 into RGB signals, to which gamma correction is applied by a gamma circuit 1010, in accordance with an output device such as a monitor. The output signal of the gamma circuit 1010 is converted by a color conversion circuit 1011 into YUV signals.
Due to downsizing and an increased number of pixels, of an image sensor, in recent years, the pixel pitch has significantly been minimized. In consequence, dark current noise and light shot noise generated by an image sensor have become conspicuous. In particular, for example, when imaging sensitivity is raised, noise conspicuously occurs. Noise can be roughly divided into dot-shaped luminance noise that occurs in a luminance signal and spot-shaped color noise that occurs in a color signal. Among these, color noise is low-frequency random noise that is a factor contributing to significant deterioration of picture quality.
To date, a method has been known in which, by utilizing a lowpass filter that passes components, of a signal, whose frequency is the same as, or lower than, that of color noise, color noise is reduced. However, because color noise has a very low frequency-response, thereby making design of a lowpass filter for reducing color noise significantly difficult, there have been problems not only in that an increased number of filter taps makes the scale of a circuit large, but also in that the effect of the lowpass filter is small. In addition, there has been a problem in that, after a color signal passes through a lowpass filter, its frequency bandwidth is narrowed and color jitters occur at the color edges of an image.
In consideration of these problems, Japanese Patent Laid-Open No. 2004-15322 discloses a distributed noise reduction method. This method is to reduce color noise, by randomly distributing low frequency noise through a random substitution method, thereby eventually converting the low frequency noise into high frequency noise, and then by processing the high frequency noise through a lowpass filter.
The distributed noise reduction method that Japanese Patent Laid-Open No. 2004-15322 proposes is suitable for color-noise reduction; however, direct application of the processing to an image-sensing apparatus caused various problems. For example, as an adverse effect of the random substitution, there is such a problem that color jitters (a phenomenon of the blurring of edge portions) occur at edge portions. Due to the fact that original color deviates to a certain hue, that is caused by saturation of an image sensor, there is such a problem that false-color jitters (blurs) occur. Also, there is such a problem that it is not possible to absorb the difference, in the extent of occurrence of noise, that depends on imaging sensitivity or that is caused by temperature variation during taking a picture.